1: \begin{abstract}

2:   In this paper, we approximate a nonautonomous neural network with infinite

3:   delay and a Heaviside signal function by neural networks with sigmoidal

4:   signal functions. We show that the solutions of the sigmoidal models converge

5:   to those of the Heaviside inclusion as the sigmoidal parameter vanishes. In

6:   addition, we prove the existence of pullback attractors in both cases, and the convergence of

7:   the attractors of the sigmoidal models to those of the Heaviside inclusion.

8:    \keywords{ Neural networks \and Differential inclusion \and

9:     Nonautonomous set-valued dynamical system \and Pullback attractor \and Upper

10:     semi convergence} \subclass{34K09 \and 34D45 \and 37B55}

11: \end{abstract}